NishanthMenon changed the topic of #openocd to: this is the place to discuss all things OpenOCD | Logs: https://libera.irclog.whitequark.org/openocd/
joconor has quit [Quit: ZNC 1.8.2 - https://znc.in]
joconor has joined #openocd
tsal has quit [Ping timeout: 248 seconds]
tsal has joined #openocd
alkane has joined #openocd
nerozero has joined #openocd
alkane has quit [Ping timeout: 268 seconds]
alkane has joined #openocd
noarb has quit [Quit: ZNC 1.8.2 - https://znc.in]
noarb has joined #openocd
alkane has quit [Ping timeout: 265 seconds]
_whitelogger has quit [Ping timeout: 260 seconds]
_whitelogger_ has joined #openocd
Hawk777 has quit [Quit: Leaving.]
zmatt has quit [Ping timeout: 246 seconds]
alkane has joined #openocd
zmatt has joined #openocd
alkane has quit [Ping timeout: 248 seconds]
alkane has joined #openocd
urja has quit [Ping timeout: 252 seconds]
urja has joined #openocd
jn has quit [Remote host closed the connection]
jn has joined #openocd
jn has quit [Remote host closed the connection]
jn has joined #openocd
jn has quit [Ping timeout: 252 seconds]
jn has joined #openocd
jn has joined #openocd
jn has quit [Ping timeout: 260 seconds]
flatmush has joined #openocd
jybz has joined #openocd
jeeebz has quit [Ping timeout: 252 seconds]
alkane has quit [Ping timeout: 265 seconds]
alkane has joined #openocd
josuah_ has joined #openocd
josuah_ has left #openocd [#openocd]
josuah has joined #openocd
<josuah> Hello! Is the OpenOCD project getting new maintainers occasionally? Enough of them? I would like to understand what makes everyone go for a custom fork instead of upstreaming.
<Haohmaru> could it be the gerrit thing?
<Haohmaru> but i'm only a user, wait for one of the devs to answer
<josuah> I'd be surprised for this to be the sole reason, but it's good to note still thanks
<Haohmaru> nah, i just accidentally found some messages in someone's github fork of openocd (was looking for a unrelated thing), and one guy asked if these changes will get "upstreamed" and the author more or less said "i tried to, registered, but it was .... and i left it"
<Haohmaru> don't remember the exact words for the "...." part ^
<Haohmaru> but also i think some of those who fork and modify openocd are just sloppy in general
<PaulFertser> josuah: what do you mean "everyone"? Is any fork packaged by any major OS yet?
<PaulFertser> josuah: yes, it would be very very good to get more maintainers but nobody volunteered recently :(
<Haohmaru> openocd is a serious program
<josuah> > sloppy in general
<PaulFertser> Becoming a maintainer is easy, one just needs to start doing good reviews for the others' code.
<josuah> yes good point, some quick-and-dirty job where OpenOCD is used to reduce the internal development effort in the first place (i.e. teams already under-staffed)
<Haohmaru> yes, that sort of thing
<josuah> PaulFertser: I meant "everyone" not as "many people" but for "anyone who do not upstream, whoever they are"
<Haohmaru> there are a lot of forks
<Haohmaru> ...which add more or less "one thing" on top of the base openocd
<Haohmaru> but you probably know better than me
<PaulFertser> josuah: well, I myself wrote a flash driver for some kinetis recently, and I sent it for review properly but it's still not reviewed and now I do not have access to hardware if changes are requested.
<Haohmaru> PaulFertser, register a 2nd accound.. FaulPertser and review it ;P~
<Haohmaru> * account
<Haohmaru> no one's gonna know
<josuah> PaulFertser: on my side, I am currently too short to afford doing reviews, but I discussed with people who are spending a bit of effort preparing their own code for reviews
<PaulFertser> josuah: it's also a bit unusual that the authors need to remind the maintainers about their code sometimes. Because indeed the priority goes to those who persist naturally, as OpenOCD has too few maintainers.
<josuah> Can people who wait their own fork reviewed speed up the process by contributing to other reviews? Hence reducing the load on reviewers?
<PaulFertser> josuah: that should work, yes, also people can come here and discuss the changes with Antonio if needed.
<PaulFertser> Or on the mailing list.
Haohmaru has quit [Quit: saionara]
<PaulFertser> josuah: do you have an example of the change that's not getting reviewed for long?
<josuah> Let me find it, good case study...
<josuah> Looks like Thomas could come to the rescue! :)
<josuah> I see such big donations for big projects, and so few going to pay staff dong the job at the end ^_^ all for the marketing/legal team (still needed, but having some devs help!)
<josuah> Is there some message for these organization who would want to put some grease in the gears of OpenOCD? As one of the project they rely on...
<PaulFertser> Hm, currently OpenOCD maintenance is purely volunteer effort, there's no reasonable way to sponsor the work I'm afraid. But money are needed for the server of course.
<PaulFertser> josuah: this BL* flash work is moving forward steadily (even if slowly) thanks to the author being competent and persistent.
<josuah> Once again, the precious resource is time more than anything :)
<josuah> Grateful for this feedback. I'll do my best to reflect it when asked, and point people at this chan to sort out practical details if needed.
<PaulFertser> Sometimes people come and part the channel too fast without giving a chance to answer them.
<gamiee> Haohmaru : actually gerrit is what I like a lot about OpenOCD. I really can't use mailing lists.
<gamiee> PaulFertser: oh, thank you!
<PaulFertser> gamiee: :) hey
<gamiee> PaulFertser , hiii :)
<borneoa___> PaulFertser: I remember someone reporting failure to setup self hosted openID. Could help collecting such info in our HACKING file?
<gamiee> also, this reminds me I need to answer Thomas, and ideally ping him on flash driver PR as well.
<josuah> gamiee: oh, hi! :D
<gamiee> hi josuah :)
<PaulFertser> borneoa___: hm, I'm not sure what would be the best way to do that. I found plenty of PHP implementations and naturally decided to look for something else, ended up using a Perl one but I had to fork it and add sqlite support and fix some stuff and it still depends on some tricky to install Perl module...
<gamiee> waaaaaaaaaaaaaaaaaaaaaait, ST-Link V3 is not allowed to be used on non STM32 MCUs?
<PaulFertser> Hahahah
<PaulFertser> Yes
<gamiee> ........
<gamiee> honestly, this is ridiculious. And I was wondering, why I am getting "unknown/unexpected STLINK status code 0x5" when it started to talk to the chip
<PaulFertser> gamiee: there's an easy way to change the firmware and I tried fixing it but I'm not a good RE and also I didn't have hardware so I never managed.
<PaulFertser> There's also https://github.com/RadioOperator/CMSIS-DAP_for_STLINK-V3MINI and probably some other projects
<gamiee> PaulFertser : oh thanks! Will take a look. Yeah, patching it would be possible, but I guess not fully worth doing it... Please, do you have any adapter to recommend which is good for SWD? (for FTDI chips it's quite confusing)
jn has joined #openocd
jn has joined #openocd
<PaulFertser> gamiee: actually with FTDI-based and resistor hack it works nicely for me.
<PaulFertser> gamiee: j-link is good obviously.
<PaulFertser> (but expensive)
<gamiee> does j-link works fine with OpenOCD?
<gamiee> for resistor hack, you need to have both two lines from FTDI connected to the SWDIO (one line without resistor and second with?)?
<PaulFertser> Yes, j-link is supported very good.
<gamiee> oh cool! Then I will try it
<PaulFertser> gamiee: if you have _bare_ FTDI board you can even spare the resistor but yes, the idea is to connect TDO via a resistor so that the target could easily override it.
<PaulFertser> OpenOCD disables drive for TDO but if you have a buffer connected to FTDI chip then it's not enough obviously.
<gamiee> I have both bare FTDI board, but also PINE64 dongle, which should have one pin (I guess TDO?) connected through resistor.
<gamiee> But I need connect both TDI and TDO (with resistor) to SWDIO, am I right?
<josuah> that also work on plenty of MCUs
<PaulFertser> gamiee: yes, see comments in ftdi/swd-resistor-hack.cfg
<josuah> including the RPi Pico, which is $5/board, cheaper than the ST Links
<gamiee> okay, good, I wasn't sure. Thanks PaulFertse
<gamiee> PaulFertser*
<PaulFertser> Is there schematics for https://pine64.com/product/usb-jtag-adapter/ ?
<PaulFertser> josuah: good thing about proper jtag adapters is that they have voltage matching circuity and termination.
<josuah> Oh right! the VRef pin...
<josuah> And level shifters...
<PaulFertser> gamiee: looks bare (plus ESD protection) to me
<PaulFertser> gamiee: TDI goes via resistor, not TDO, sorry.
<PaulFertser> TDI is driven by adapter, it's input for the target.
<PaulFertser> So if there's no buffering you can just connect both TDO and TDI to SWDIO and that's it.
<PaulFertser> (and source the resistor-hack config)
<gamiee> yeah, no buffering + the resistor is available with another pin exposed from adapter
<PaulFertser> That should work, yes. With the resistor or without, doesn't matter.
<gamiee> great! Thanks :)
noarb has quit [Quit: ZNC 1.8.2 - https://znc.in]
<karlp> gamiee: nxp's mculink is cmsis-dap, with SWO support, HS USB, and has a nice util uart on the side, available from mouser/digikey, and isn't part locked. can recommend.
<karlp> runs around ~15€ from major disties
noarb has joined #openocd
<gamiee> karlp : oh cool! Thanks for suggestion
noarb has quit [Quit: ZNC 1.8.2 - https://znc.in]
noarb has joined #openocd
<gamiee> PaulFertser : the SWD hack with FTDI worked, thanks!
<PaulFertser> gamiee: yay :)
nerozero has quit [Ping timeout: 252 seconds]
russ has joined #openocd
* karlp cheers
russ has quit [Ping timeout: 260 seconds]